실험을 한 번도 하지 않고 컴퓨터만으로 신약 후보 하나를 고르는 절차가 처음부터 끝까지 한 편에 들어 있다. 표적 Ror1 은 정상 성인 조직에는 거의 없고 암에서만 다시 켜지는 단백질이라, 지금도 항체와 CAR-T 로 임상에 올라 있다. 다만 이 논문은 원자료도 코드도 남기지 않았다 — 그래서 인실리코 결과 중 무엇을 믿고 무엇을 다시 확인해야 하는지를 배우기에 오히려 좋은 호다.
You are at home in a shell and in data structures; it is the biology that stops you.
This order is computed from what the paper needs and how the courses depend on one another — change the paper and the order changes with it.
768 MB · 1 core. The four editions and a cut-down notebook — real scanpy code on a 2,000-cell subsample, done in 30 seconds.
8 GB · 4 cores, the scanpy stack installed and the raw data mounted at /data. It costs, so it sleeps after 10 idle minutes — your files stay.
| Secondary and tertiary structure analysis | Protein Structure Prediction, First Steps |
| Structure absent from RCSB PDB and SCOP | Protein Structure Prediction, First Steps |
| PROMOTIF and GOR4 secondary structure prediction | Protein Structure Prediction, First Steps |
| Fig. 2 — Template search using BLASTp | Protein Structure Prediction, First Steps |
| 3D structure prediction (CPHmodels, Phyre2, PS2v2, RaptorX, Modeller, I-TASSER) | Protein Structure Prediction, First Steps |
| Fig. 3a-c — ProSA Z-score and per-residue energy | Protein Structure Prediction, First Steps |
| SAVES server — 97.16% of residues 3D-1D >= 0.2 | Protein Structure Prediction, First Steps |
| PROCHECK — 67.4 / 27.3 / 3.8 / 1.6 % | Protein Structure Prediction, First Steps |
| Schrodinger Protein Preparation Wizard | Protein Structure Prediction, First Steps |
| Fig. 5 — Domain analysis by CD-Search (Pkc_like superfamily) | Protein Structure Prediction, First Steps |
| Active site prediction — DogSiteScorer and SiteMap | Protein Structure Prediction, First Steps |
| Table 1 — Active site analysis using DogSiteScorer | Protein Structure Prediction, First Steps |
| Fig. 1 — Secondary structure analysis of Ror1 | Protein Structure Prediction, First Steps |
| Fig. 9 — Ror1 signalling pathway (IL-6/STAT3/Wnt5a/PI3K-AKT) | Protein Structure Prediction, First Steps |
| Protein-Ligand Docking and MD simulation | Docking and Molecular Dynamics, First Steps |
| Ligand preparation — ZINC database, 45,000 unique ligands | Docking and Molecular Dynamics, First Steps |
| LigPrep optimisation | Docking and Molecular Dynamics, First Steps |
| Glide HTVS screening, cut-off -7 kcal/mol | Docking and Molecular Dynamics, First Steps |
| Fig. 6 — Drug-likeness screening parameters (FAF-Drugs3) | Docking and Molecular Dynamics, First Steps |
| AutoDock Vina rescoring, |affinity| >= 10.9 | Docking and Molecular Dynamics, First Steps |
| Table 2 — Glide scores of the best 10 docked ligands | Docking and Molecular Dynamics, First Steps |
| Fig. 7a-b — Hydrogen bond interactions, protein + ligand 15 | Docking and Molecular Dynamics, First Steps |
| ADMET properties of DB03208 (PGG) | Docking and Molecular Dynamics, First Steps |
| GROMACS system setup — SPC/E water, AMBER99SB-ILDN, 3 Cl- | Docking and Molecular Dynamics, First Steps |
| Energy minimisation, 0.1 ns, emtol 10.0 | Docking and Molecular Dynamics, First Steps |
| Fig. 8a — RMSD of protein, ligand and complex | Docking and Molecular Dynamics, First Steps |
| Fig. 8b-c — RMSF of protein and ligand | Docking and Molecular Dynamics, First Steps |
| High-throughput MM-PBSA by g_mmpbsa (top 300 ligands) | Docking and Molecular Dynamics, First Steps |
| Molecular basis — polyphenols, hydrogen bonding, logP | ⚗️ Chemistry for Bioinformatics, First Steps |
| In-silico drug discovery workflow end to end | 🎯 Insilico Medicine, First Steps |
| Target selection — Ror1 as a cancer target | Target Discovery, First Steps |
| ADMET / absorption, BBB, water solubility | 💊 Pharmacokinetics, First Steps |
Let's Read the Landmark Papers